My take on why indeed the study of consciousness may not be as simple

[Gate2501]

Yeah, you should make 'em look foolish by coming up with a neat definition, possibly involving the word "just".

Right, I will use the word "just" less.

How are you not engaging in Loki's Wager when you deliberately attempt to make consciousness impossible to define?

If it is actually fundamentally impossible to define, wth are you even arguing about?

 

[rocketdodger]

Is that really what he said?

~~ Paul

Not exactly (I think). I will elaborate later, I am leaving work now.

 

[Robin]

Sorry, I don't follow.

How am I looking at two hands, the information for which is light years away from each other with no physical pathway?

You deciding about yourself is one thing. The problem is that we haven't determined how we are going to decide if a computer (in one piece or far-flung) is conscious.

I am not really concerned with that.

As far as I am concerned if a computer can pass the Turing Test then it understands and it thinks. Whether it is conscious (when observed objectively) is a pseudo problem.

I have specifically framed the question in terms of whether this moment I am experiencing right now could be a computer running an algorithm.

Even concerning yourself, you have the problem that you may be a zombie set up to think she is conscious.

In which case "real" consciousness is something I am not concerned with. It is some unrelated and unknowable thing.

What I am asking about then is this illusion I have.

So let's say we teleport a copy of your brain every millisecond for a few seconds. Does the array of brains embody consciousness?

As long as each brain has sufficient connected information for an experiencable time slice then no problem.

There is no problem then of taking these brains light years apart - I could be that because there is sufficient information with a physical pathway in each slice of time.

As I said I cannot tell at any moment if this was the first moment I existed and will be the last.

The problem with Run4 is that there is not even enough information for a single time slice in any one probe.

 

[Robin]

Robin is saying they are the same (in that each run produces a unique instance of the consciousness) because the algorithms are the same.

No, I said they are the same because exactly the same set of instructions were run in exactly the same order physically on the box.

The only thing that changed were the context switching and memory handling routines in between and there is no overall nouse in the system to know that the 5 retrieved in step n was not the same 5 saved in step n-5.

 

[Malerin]

Do you think you might be a p-zombie?

Maybe. I couldn't tell if I were, anyway.

And the thread reaches its absurdity saturation point.

 

[AkuManiMani]

And the thread reaches its absurdity saturation point.

The Pit of Fail has no bottom, sir.

 

[Malerin]

The Pit of Fail has no bottom, sir.

Indeed.

 

[Robin]

Your example program was also unduly complicated to go through in a forum discussion, so let me fix that.

Wow, your "simplification" was much more complicated than my original example. How does it differ?

The processor cycle, let's say, is step Y. X is one of the rest of the steps. Y isn't in total isolation of X. What happens at Y should depend on X. If it doesn't, that means that memory is failing to supply its black box guarantees that we rely on in order to run the algorithm.

And yes, that is what I'm claiming.

So are you or are you not claiming that there is something in the system what will be able to tell that the black box has not lived up to it's guarantee, so that this processor cycle should somehow produce some different result, even though the value of the datum that was stored was the same as the one that was retrieved?

Now I have to say "something" and "some different result" because I don't know what it is that you and rocketdodger are referring to here. I can't understand what you are trying to say.

As far as I know when the CPU processes a particular instruction it has no idea where the data came from beyond the memory location specified and no idea where it is going to go beyond the memory location specified.

It does not know how that datum got into that particular memory location and it does not need to know.

Furthermore, I thought you said you were showing what the difference is.

I was??? Where? I thought that you and rocketdodger were claiming that those instructions would produce a different result in Run3 as they did in Run1.

If not, then no problem.

But regardless, part 2 being completely different than part 1, and this "showing a difference" thing somehow mutating within your post into "same thing", and "set" being the same as "cycle", etc... I cannot make a lick of sense out of the aforementioned quoted text. Could you please help?

What on earth are you talking about here? What is part 1 and part 2? What is before and after the dash?

Are you referring to something I said here? I have no idea what you are even asking.

It is simple

Run 1

Instruction 1 stores A
Black box
Instruction 2 retreives A does calculation and stores B
Black box
Instruction 3 retreives B

Run 3
Instruction 1 stores A
Black box 2
Instruction 2 retreives A does calculation and stores B
Black box 2
Instruction 3 retreives B


Now so long as instruction 2 gets back the same value that was stored in instruction 1 and instruction 3 gets back the same value that was stored in instruction 2, will the process represented by {Instruction 1, Instruction 2, Instruction 3} behave the same way in Run 3 as it did in Run 1?

Why is that such a difficult question to answer?

 

[Robin]

Robin: So I am running an algorithm. The current step is to add 27 and 95. What does that tell me about the last step?
PixyMisa: It tells you absolutely nothing about the last step. Why should it?

OK, we have established that you are unable to tell what algorithm was or even what the last step was.

But, without this information, are you able to add 27 to 95?

No, that's trivial. We just have to change the algorithm so that the current step is to examine the previous step.

Once it has examined what the previous step was, does the CPU now know what the previous step was?

I think you would have to change the algorithm to examine what the current step was so when it completes it will be the previous step.

So in that case the previous step will always be to examine the current step.

Will this information help the CPU carry out the new current step?

 

[Dancing David]

And the thread reaches its absurdity saturation point.

How would you know you aren't a p-zombie. A p-zombie would believe it was conscious and would have experiences that made it believe it was conscious. That is the way it is contructed, it has everything that can be described as consciousness in terms of behavior, so it has the belief it is conscious, it has events that make it feel it is conscious, it just isn't.

That is the problem with a p-zombie.

ETA: How could anyone tell it was a p-zombie and not conscious.

 

[Robin]

How would you know you aren't a p-zombie. A p-zombie would believe it was conscious and would have experiences that made it believe it was conscious. That is the way it is contructed, it has everything that can be described as consciousness in terms of behavior, so it has the belief it is conscious, it has events that make it feel it is conscious, it just isn't.

That is the problem with a p-zombie.

ETA: How could anyone tell it was a p-zombie and not conscious.

Make it feel that it is conscious?

As I said to Paul, if I am not really conscious, just being tricked into feeling I am conscious then consciousness is an irrelevant unknowable thing.

So it is this feeling that I am conscious that I have right now that I am calling consciousness.

 

[rocketdodger]

Is that really what he said?

~~ Paul

Alright, I can explain now.

I said (or at least I am saying now) that only the first run produces a unique consciousness, and that all other runs are simply further instances of this same consciousness.

This is assuming, of course, that Run1 is the first instance of the algorithm in the universe. If it isn't, then Run1 is also merely another instance of the same consciousness.

 

[Robin]

No, that's trivial. We just have to change the algorithm so that the current step is to examine the previous step.

Once it has examined what the previous step was, does the CPU now know what the previous step was?

I think you would have to change the algorithm to examine what the current step was so when it completes it will be the previous step.

So in that case the previous step will always be to examine the current step.

Will this information help the CPU carry out the new current step?

Also, how does the CPU know that the information it receives from (IP-1) is, in fact, accurate?

 

[Robin]

Alright, I can explain now.

I said (or at least I am saying now) that only the first run produces a unique consciousness, and that all other runs are simply further instances of this same consciousness.

This is assuming, of course, that Run1 is the first instance of the algorithm in the universe. If it isn't, then Run1 is also merely another instance of the same consciousness.

Here is the post I was referring to:

And so according to the theory a real unified human consiousness would result across many light years with no communication whatsoever between the components creating this consciousness.

No -- you haven't been paying attention.

The consciousness occurs when you program the devices, not when they execute their instructions.

In other words, the "recheck" doesn't mean anything, only the initial run. And everything else is simply a remapping of the initial run. The initial run is the consciousness.

 

[rocketdodger]

I said nothing of the sort.

You are attributing to me your own intuition that an algorithm is somehow a thing.

I don't consider that the program was a single instance even in the original run except in a purely abstract sense.

In Run1 all that happened was that a bunch of separate operations linked by the fact that they are related by some other program by an index in the process table and that they will write the the same piece of virtual memory and that someone has previously taken care that these operations in this particular order will result in some meaningful calculation.

Even in Run1 the registers could have been overwritten hundreds of times between any two steps in the program, and the memory it writes to could have been overwritten many times. The program state might have been saved to disk at any point, moved to a different computer and the original hardware crushed to dust, then the program started up again and the algorithm would have continued just as though it had been run on a dedicated CPU.

But an algorithm is a thing. What else could it be? Are you claiming that algorithms are somehow nonphysical, that they are floating in the void?

Look, an algorithm is simply a step by step breakdown of the behavior of a physical system.

And if you break down the steps of Run3, you see many steps of Run2 included. How could you not, given that the states of Run3 are produced during Run2?

When that program step is loaded to the CPU it is run with the specific data items and register state it is given and there is nothing on the system that knows that between ST A (fefa) in one step and LD A (fefa) in another that same value for (fefa) was restored from the wrong place.

What is it that knows this?

The algorithm could care less. Would you care if one of your neurons was replaced with another one? Absolutely not.

But we know. And what you are failing to realize, for some reason or another, is that a consciousness -- you, for example -- wouldn't know whether it was the same one as before, or different, or if there were multiple instances of you, or whatever. All you have access to is the private behavior of the instance of you that is ... you.

 

[rocketdodger]

Here is the post I was referring to:

Yeah I meant what I said in my latest post about that. Sorry if it was worded badly originally.

 

[AkuManiMani]

How would you know you aren't a p-zombie. A p-zombie would believe it was conscious and would have experiences that made it believe it was conscious. That is the way it is contructed, it has everything that can be described as consciousness in terms of behavior, so it has the belief it is conscious, it has events that make it feel it is conscious, it just isn't.

That is the problem with a p-zombie.

DD, what you just said reveals that not only do you not understand the concept of p-zombies, but that you also do not understand what consciousness means. For an entity to be a p-zombie means that, by definition, it cannot feel or believe anything.

 

[Robin]

But an algorithm is a thing.

What sort of a thing?

Look, an algorithm is simply a step by step breakdown of the behavior of a physical system.

Not necessarily, you can have a valid algorithm that could never even in principle run on a physical system such as the Seive of Eratosthenes that I showed earlier.

And if you break down the steps of Run3, you see many steps of Run2 included.

You see all the steps from Run2 and only the steps from Run2. They are identical.

How could you not, given that the states of Run3 are produced during Run2?

Of course you could not. So why would this case be different?

The algorithm could care less. Would you care if one of your neurons was replaced with another one? Absolutely not.

And yet I would still be conscious, so why wouldn't Run3?

But we know. And what you are failing to realize, for some reason or another, is that a consciousness -- you, for example -- wouldn't know whether it was the same one as before, or different, or if there were multiple instances of you, or whatever.

What on earth makes you think I am failing to realise that???

I have stipulated several times that I could not know whether or not this moment was my first and will be my last.

But Run1 produces consciousness. Does Run2 produce consciousness? I am still not sure of your position on this.

If they both produce the same concsciousness then consciousness is non-temporal, therefore non-physical.

So presumably Run2 produces an identical consciousness but later in physical time.

But if Run2 produces consciousness within physical time then so should Run3.

 

[PixyMisa]

How would you know you aren't a p-zombie. A p-zombie would believe it was conscious and would have experiences that made it believe it was conscious. That is the way it is contructed, it has everything that can be described as consciousness in terms of behavior, so it has the belief it is conscious, it has events that make it feel it is conscious, it just isn't.

That is the problem with a p-zombie.

ETA: How could anyone tell it was a p-zombie and not conscious.

The concept of a p-zombie is incoherent under materialism, and also under some forms of idealism. So the way you can tell that you're not a p-zombie is the same as the way that you can tell you're not a square circle.

 

[PixyMisa]

Alright, I can explain now.

I said (or at least I am saying now) that only the first run produces a unique consciousness, and that all other runs are simply further instances of this same consciousness.

This is assuming, of course, that Run1 is the first instance of the algorithm in the universe. If it isn't, then Run1 is also merely another instance of the same consciousness.

Right. If run 1 is an algorithm that is actually interacting with the Universe - or with something, anything, external, then replaying from a recording it replays an identical consciousness. Whether it's the same consciousness depends on how you define "consciousness", and, indeed, "same". Saying that the other runs produce another instance of that consciousness is certainly valid.